In South Africa more than 11% of the population are infected with HIV-1 so there is a clear need for a prophylactic HIV-1 vaccine. The focus of this application is the use of Mycobacterium bovis bacille Calmette-Gu[unreadable]rin (BCG) as an HIV vaccine vector. BCG - which is better known as the tuberculosis vaccine - is widely used to immunize infants, has strong adjuvant potential in man and animals, can be administered orally, is inexpensive to produce, is heat stable, elicits long-lasting cellular immune response and has a very low rate of complications. These characteristics make BCG a very attractive vehicle for recombinant vaccines. This vaccine is likely to induce good cellular immune responses but not neutralising antibodies. Despite these advantages there are also problems expressing viral antigens in BCG, which result in low immunogenicity and genetic instability. The hypothesis being tested in this application is that immunogenicity and stability of recombinant BCG (rBCG) expressing HIV antigens can be improved either through modification of the HIV genes or by regulation of HIV gene expression. The aims of the proposal are: To modify HIV genes to increase the stability of rBCG and increase immunogenicity. To develop systems which allow for down regulation of expression of HIV genes in rBCG and up regulation in animals. To assess the immunogenicity of rBCG expressing HIV genes in combination with matched virus like particles (VLP) and modified vaccinia virus Ankara (MVA) based HIV candidate vaccines initially in mice and then in non-human primates. The vaccine will be designed for South Africa where HIV-1 subtype C is the dominant circulating subtype. Genes were selected from recently transmitted HIV and closest to the South African consensus sequence. As it is likely that a vaccine will be more effective with more genes, this project will initially focus on gag and reverse transcriptase but will then move on to include the gene encoding HIV-1 subtype C gp120.